Normally, during using the digital video camera, making an example of the digital camera for capturing still images, the user usually will hold the camera upright for picturing, that is, the horizontal width is larger than the vertical height to be conformed with the physiological feature of human using both eyes to view the external objects for picturing. However, because sometimes the target object to be shot has larger height than width, or the view to be found has an angle with larger height than width, the digital camera will be rotated from the horizontal orientation status to the vertical orientation status for picturing. Due to the configuration of sensing elements in the digital camera, the captured image will become the horizontal orientation status when directly viewing using an ordinary computer image playing software. Therefore, the user will need to employ the counterclockwise or clockwise rotation function to rotate the pictures one by one and then to store, so that it is very inconvenient for the user to read and view the images.
In the prior art for image rotation, there are some related control techniques being developed. As for the patent technique, the R.O.C. patent publication no. I276042 discloses an automatic image rotation device and the implementation method thereof. In this patent, the automatic image rotation device comprises a load unit, a memory unit, a micro-controller, and a display unit; wherein, the micro-controller detects the image features in the image captured by the image capturing device and loaded by the load unit according to the feature recognition algorithm stored in the memory unit, and determines the image features in the image according to the feature recognition algorithm.
When the feature recognition algorithm determines the image features in the image are not upright, the micro-controller will rotate the images until the image features in the image becoming upright. Furthermore, the micro-controller will display the pictures on the display unit after rotating the images before rotating the images for preview. The automatic image rotation method includes a load step for loading an image, a detection step for detecting at least one image feature in the image, and a determination step for determining if the image features in the image are upright, and, if not, a rotation step for rotating the image until the image features in the image becoming upright.
However, in order to correctly observe the target image, it is necessary to rotate the electronic product for picturing during shooting to be conformed with the angular orientation of the target to be shot, or, in the previous case for the automatic image rotation device and the implementation method thereof, it employs the feature recognition algorithm stored in the memory unit to determine if there is any image feature contained in the target image, and then employs the determination step of the micro-controller to execute the rotation step for rotating the target image until the image features in the image are upright.
No matter rotating the electronic product for picturing to obtain the suitable shooting orientation angle or conducting multiple rotation on the image features contained in the target image with the micro-controller, it still needs to rotate the stored target images using the image features. And, the above-mentioned image features should at least include a human face feature, such as eyes, eyebrows, nostril, lips, ears and hairs. As for other images not including human face features, such as objects, devices and scenes, it could not be determined to execute the image rotation.